Incremental heating is accomplished using a fully automated, 75 W, 980 nanometer wave length diode laser system.
Tantalum or niobium-encapsulated sample
Tantalum or niobium-encapsulated samples are heated in two modes: (1) resting upon a stainless steel planchet; or (2) suspended on thermocouple wire extended from one of seven available ports of a sperical octagon. The diameter of a collimated laser beam is adjusted to match the diameter of 1 mm Nb packets used for single crystals or 3mm diameter, Ta foil packets used for multi-grain aggregates. The 30 cc laser chamber is rigidly mounted. An actuated Zaber x-y stage positions the laser. A 1.9" viewable sapphire window permits sufficiently widely separated samples to avoid beam overlap.
45 Well He Pan
21 Well AR Pan
Trays designed to hold up to forty five 1 mm packets spaced at 4 mm can be loaded for an automated run. Alternative trays hold 21 of the larger 3 mm packets spaced at 6mm. Heating is monitored with a firewire IR Guppy CCD camera mounted on a Navitar camera and a 250-1650 C, 0.85 mm spot optical pyrometer mounted on an actuated Zaber mirror mount. Laser output power is controlled either: (1) directly (e.g.., Current Mode) via internal current modulation in the OSTECH diode laser controller; or (2) via external modulation (e.g., PID mode) based upon thermocouple feedback or optical pyrometer response.
The temperature accuracy of optical pyrometry measurements can be checked by monitoring release of 40Ar upon melting from a disc of alloy 1100 (ultra corrosion resistant) aluminum that has been enclosed within a standard 3 mm diameter Ta foil packet in the same manner as for unknown silicate samples. This approach to temperature calibration is possible because Ar is trapped in the alloy 1100 manufacturing process due to the use of argon as purge gas to remove hydrogen gas. We have determined that a 3mm diameter x 0.05 mm thick disc of Al foil releases ca. 2.5x10-14 mol of 40Ar as it melts. Based upon incremental heating of Ta-encapsulated Al foil, the current temperature accuracy of optical pyrometry measurements made from the Ta surface is +/- 18oC when 10 samples are run in sucession. This variability results from mis-allignment of the pyrometer sample region and the laser image upon the sample when moving between samples. When the pyrometer is alligned for each new sample, the temperature accuracy is improved considerably.
Confirming the accuracy of the diode laser pyrometer measurements with from Arrhenius data calculated from incremental heating of single air abraded spheres of Kalsilite glass. Heating steps are three minutes. The temperature of the step is calculated by weighting each 1 second temperature integration by an exponential function that simulates the diffusion behavior of silicates: Exp[-25,000/RT]. In this way, only the high-temperature steps near the set point actually matter. The duration of the heating step is calculated from the time that the temperature is within 2 siggma of the weighted mean temperature as calculated above.
Temperature calibration is performed using the same procedures used to heat unknown samples and is based upon the melting point of Al foil and the Arrhenius behavior of kalsilite glass. For diffusion measurements, reproducible sample heating is ensured by using routines to center the sample beneath the laser beam and the optical pyrometer upon the sample.
